Published image generation rate
Meta lists Muse Image at $0.01 per successfully generated and returned image. The price page says prompt length, reasoning strength and built-in generation tools do not change that flat rate. A request returning several images is charged for the number returned; failed or safety-filtered images are not counted under that stated rule.
This is the direct Meta Model API tariff as checked on 17 October 2026. It is not a consumer-app subscription price or a third-party reseller quote.
Illustrative image budgets
| Returned images | Calculation | API generation estimate |
|---|---|---|
| 100 | 100 × $0.01 | $1.00 |
| 1,000 | 1,000 × $0.01 | $10.00 |
| 10,000 | 10,000 × $0.01 | $100.00 |
These are chosen quantities, not observed invoices or a throughput claim. If you ask for three alternatives and then generate two revisions, budget for all five returned pictures. Your product may display only one final selection while the workflow produced several billable results.
Keep storage, delivery, other model calls and applicable taxes outside this narrow estimate. A planning spreadsheet should show those costs separately instead of hiding them inside an invented blended image rate.
Generation, editing and understanding
Use the image-generation documentation for creating or editing an image with muse-image-1.0. The documentation describes both Responses-based workflows and image endpoints. A text-output model answering a question about a screenshot is a different operation.
Likewise, video input does not imply video generation. Before estimating a multimedia product, identify whether you need text analysis, image output, transcription or segmentation. A model’s supported input modalities do not determine the billing unit of every output task.
Track cost through the product workflow
- Count successfully returned images by request and user task.
- Include alternatives and revisions even when they are not displayed.
- Separate provider-reported image billing from token usage shown for reference.
- Check errors and retry behavior to avoid producing unwanted duplicate results.
- Compare actual invoices against the returned-image count before scaling.
For Wonder’s token arithmetic, return to API pricing. For choosing between text and image-output model routes, use the model selection guide. No images were generated to prepare this article.
Sources & evidence
Editorial check: 17 October 2026. Product facts come from the following primary sources; planning checklists and examples are our explanations. No paid inference, model benchmark or regional checkout was run for this guide.